Dataset Information

When the Eyes No Longer Lead: Familiarity and Length Effects on Eye-Voice Span.

ABSTRACT: During oral reading, the eyes tend to be ahead of the voice (eye-voice span, EVS). It has been hypothesized that the extent to which this happens depends on the automaticity of reading processes, namely on the speed of print-to-sound conversion. We tested whether EVS is affected by another automaticity component - immunity from interference. To that end, we manipulated word familiarity (high-frequency, low-frequency, and pseudowords, PW) and word length as proxies of immunity from interference, and we used linear mixed effects models to measure the effects of both variables on the time interval at which readers do parallel processing by gazing at word N + 1 while not having articulated word N yet (offset EVS). Parallel processing was enhanced by automaticity, as shown by familiarity × length interactions on offset EVS, and it was impeded by lack of automaticity, as shown by the transformation of offset EVS into voice-eye span (voice ahead of the offset of the eyes) in PWs. The relation between parallel processing and automaticity was strengthened by the fact that offset EVS predicted reading velocity. Our findings contribute to understand how the offset EVS, an index that is obtained in oral reading, may tap into different components of automaticity that underlie reading ability, oral or silent. In addition, we compared the duration of the offset EVS with the average reference duration of stages in word production, and we saw that the offset EVS may accommodate for more than the articulatory programming stage of word N.

SUBMITTER: Silva S

PROVIDER: S-EPMC5089997 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundIndividuals with autism spectrum disorder (ASD) and their parents demonstrate impaired performance in rapid automatized naming (RAN), a task that recruits a variety of linguistic and executive processes. Though the basic processes that contribute to RAN differences remain unclear, eye-voice relationships, as measured through eye tracking, can provide insight into cognitive and perceptual processes contributing to RAN performance. For example, in RAN, eye-voice span (EVS), the distance ahead the eyes are when articulation of a target item's label begins, is an indirect measure of automaticity of the processes underlying RAN. The primary objective of this study was to investigate automaticity in naming processes, as indexed by EVS during RAN. The secondary objective was to characterize RAN difficulties in individuals with ASD and their siblings.MethodsParticipants (aged 15-33 years) included 21 individuals with ASD, 23 siblings of individuals with ASD, and 24 control subjects, group-matched on chronological age. Naming time, frequency of errors, and EVS were measured during a RAN task and compared across groups.ResultsA stepwise pattern of RAN performance was observed, with individuals with ASD demonstrating the slowest naming across all RAN conditions, controls demonstrating the fastest naming, and siblings demonstrating intermediate performance. Individuals with ASD exhibited smaller EVSs than controls on all RAN conditions, and siblings exhibited smaller EVSs during number naming (the most highly automatized type of naming). EVSs were correlated with naming times in controls only, and only in the more automatized conditions.ConclusionsThese results suggest that reduced automaticity in the component processes of RAN may underpin differences in individuals with ASD and their siblings. These findings also provide further support that RAN abilities are impacted by genetic liability to ASD. This study has important implications for understanding the underlying skills contributing to language-related deficits in ASD.

Project description:BackgroundWhat can theories regarding memory-related gaze preference contribute to the field of deception detection? While abundant research has examined the ability to detect concealed information through physiological responses, only recently has the scientific community started to explore how eye tracking can be utilized for that purpose. However, previous attempts to detect deception through eye movements have led to relatively low detection ability in comparison to physiological measures. In the current study, we demonstrate that the modulation of gaze behavior by familiarity, changes considerably when participants perform a visual detection task in comparison to a short-term memory task (that was used in a previous study). Thus, we highlight the importance of theory-based selection of task demands for improving the ability to detect concealed information using eye-movement measures.ResultsDuring visual exploration of four faces (some familiar and some unfamiliar) gaze was allocated preferably on familiar faces, manifested by more fixations. However, this preference tendency vanished once participants were instructed to convey countermeasures and conceal their familiarity by deploying gaze equally to all faces. This gaze behavior during the visual detection task differed significantly from the one observed during a short-term memory task used in a previous study in which a preference towards unfamiliar faces was evident even when countermeasures were applied.ConclusionsDifferent tasks elicit different patterns of gaze behavior towards familiar and unfamiliar faces. Moreover, the ability to voluntarily control gaze behavior is tightly related to task demands. Adequate ability to control gaze was observed in the current visual detection task when memorizing the faces was not required for a successful accomplishment of the task. Thus, applied settings would benefit from a short-term memory task which is much more robust to countermeasure efforts. Beyond shedding light on theories of gaze preference, these findings provide a backbone for future research in the field of deception detection via eye movements.

Dataset Information

When the Eyes No Longer Lead: Familiarity and Length Effects on Eye-Voice Span.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets